[SUPPORT] Schema evolution using DataSource and HiveSyncTool hudi 0.15.0

**Describe the problem you faced**
hello i try to test several schema evolution usecases using hudi 0.15 and spark3.5 using hms 4
first test: Adding column in PG --> debezium / schema registry ok --> hudi (MOR) hivesyncTool KO
```
org.apache.hudi.hive.HiveSyncTool.syncHoodieTable(HiveSyncTool.java:250)
at org.apache.hudi.hive.HiveSyncTool.doSync(HiveSyncTool.java:193)
at org.apache.hudi.hive.HiveSyncTool.syncHoodieTable(HiveSyncTool.java:167)
... 69 more
Caused by: InvalidOperationException(message:The following columns have types incompatible with the existing columns in their respective positions :
username)
```
2. type promotion following this doc: https://hudi.apache.org/docs/schema_evolution/#type-promotions in PG , Double to String ok --> debezium / schema registy ok --> hudi (MOR) hivesyncTool KO
`Caused by: org.apache.hudi.hive.HoodieHiveSyncException: Could not convert field Type from DOUBLE to string for field salary`
3. DROP column in PG --> debezium / schema registy ok --> hudi (MOR) hivesyncTool job not failed but i see always the column but with null values for newest inserts
my configuration is :
```
"hoodie.datasource.hive_sync.ignore_exceptions" -> "true",
"hoodie.write.set.null.for.missing.columns" -> "true",
"hoodie.schema.on.read.enable" -> "true"
hive.metastore.disallow.incompatible.col.type.changes"-> "false"
```
when i drop tables (ro / rt ) and i restart my job is creating new tables correctly, but is not production way to handle schemas evolutions

**To Reproduce**

Steps to reproduce the behavior:

1. on the PG source side:
`cdc_hudi=> ALTER TABLE employees ADD COLUMN test_str VARCHAR ;
ALTER TABLE
cdc_hudi=> INSERT INTO employees (name, department, username, test_str) VALUES ('armel011', 'Engineering', 'arm23220', 'teststr');
INSERT 0 1`
2.debezium / schema registry ok ( latest version contains added columns)

3.restart spark job

**Expected behavior**

Case 1. add column
Case 2: Rename column
Case 3. change datatype double to string
Case 4. drop column

**Environment Description**

* Hudi version : 0.15.0

* Spark version : 3.5.1

* Hive version : 4.0.0

* Hadoop version : 3.4

* Storage (HDFS/S3/GCS..) : S3

* Running on Docker? (yes/no) : no

**Stacktrace**

```Caused by: org.apache.hudi.exception.HoodieException: Got runtime exception when hive syncing employees
at org.apache.hudi.hive.HiveSyncTool.syncHoodieTable(HiveSyncTool.java:170)
at org.apache.hudi.sync.common.util.SyncUtilHelpers.runHoodieMetaSync(SyncUtilHelpers.java:79)
... 68 more
Caused by: org.apache.hudi.hive.HoodieHiveSyncException: Failed to update table for employees_ro
at org.apache.hudi.hive.ddl.HMSDDLExecutor.updateTableDefinition(HMSDDLExecutor.java:162)
at org.apache.hudi.hive.HoodieHiveSyncClient.updateTableSchema(HoodieHiveSyncClient.java:205)
at org.apache.hudi.hive.HiveSyncTool.syncSchema(HiveSyncTool.java:347)
at org.apache.hudi.hive.HiveSyncTool.syncHoodieTable(HiveSyncTool.java:250)
at org.apache.hudi.hive.HiveSyncTool.doSync(HiveSyncTool.java:193)
at org.apache.hudi.hive.HiveSyncTool.syncHoodieTable(HiveSyncTool.java:167)
... 69 more
Caused by: InvalidOperationException(message:The following columns have types incompatible with the existing columns in their respective positions :
test_str)
at org.apache.hadoop.hive.metastore.api.ThriftHiveMetastore$alter_table_with_environment_context_result$alter_table_with_environment_context_resultStandardScheme.read(ThriftHiveMetastore.java:59744)
at org.apache.hadoop.hive.metastore.api.ThriftHiveMetastore$alter_table_with_environment_context_result$alter_table_with_environment_context_resultStandardScheme.read(ThriftHiveMetastore.java:59730)
at org.apache.hadoop.hive.metastore.api.ThriftHiveMetastore$alter_table_with_environment_context_result.read(ThriftHiveMetastore.java:59672)
at org.apache.thrift.TServiceClient.receiveBase(TServiceClient.java:88)
at org.apache.hadoop.hive.metastore.api.ThriftHiveMetastore$Client.recv_alter_table_with_environment_context(ThriftHiveMetastore.java:1693)
at org.apache.hadoop.hive.metastore.api.ThriftHiveMetastore$Client.alter_table_with_environment_context(ThriftHiveMetastore.java:1677)
at org.apache.hadoop.hive.metastore.HiveMetaStoreClient.alter_table_with_environmentContext(HiveMetaStoreClient.java:373)
at org.apache.hadoop.hive.ql.metadata.SessionHiveMetaStoreClient.alter_table_with_environmentContext(SessionHiveMetaStoreClient.java:322)
at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62)
at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
at java.lang.reflect.Method.invoke(Method.java:498)
at org.apache.hadoop.hive.metastore.RetryingMetaStoreClient.invoke(RetryingMetaStoreClient.java:173)```

Start with the failing path in HiveSyncTool.java, then read HiveSyncTool.syncSchema, HoodieHiveSyncClient.updateTableSchema, and HMSDDLExecutor.updateTableDefinition. Reproduce the reported add-column and type-promotion cases with PostgreSQL CDC, Spark, and Hive, then establish whether Hive sync handles the four listed schema changes without incompatible-type errors or stale columns.

Written by the indexing model from the issue text.